We propose to continue our investigations into "Magnetic Resonance Fluoroscopy." During years 04-06 of this grant we integrated ultrashort TR (<10 msec) data acquisition with high speed reconstruction to demonstrate realtime MR image formation. The three major projects proposed for the next grant period are: 1. Realtime MR Fluoroscopy. We will continue our investigations in continuous high speed realtime MR imaging in three sub-projects: (a) Technique Optimization of the data acquisition including a quantitative study of various tradeoffs, (b) development of Instrumentation for interactive MRI, and (c) studies of several Applications of MR Fluoroscopy. As a result of our work in Years 04-06 we have identified two areas of investigation distinct from but related to MR Fluoroscopy: 2. Snapshot FLASH MRI. Rather than use a continuous short TR sequence, we will append a burst of short TR repetitions to measure the contrast created by a preparatory pulse sequence such as inversion recovery. The goal is to generate diagnostic images with acquisition times less than 1000 msec. Sub-projects are: (a) T1-Weighted Imaging, (b) T2-Weighted Imaging, and (c) Instrumentation. 3. Intra-Scan Modification of Scan Parameters. In our investigations in Years 04-06 we devised the technique of variable echo times. We will apply this to standard spin-echo MRI. This can potentially provide edge enhancement, improved SNR, or reduced scan time with negligible reduction in contrast. Sub-projects are: (a) Variable Spin-Echo Imaging (VARSE), and (b) Variable Inversion Recovery Imaging.